<div dir="auto">Thanks for the valuable perspective and insights!</div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">El sáb., 8 de ene. de 2022 2:07 p. m., Kevin Venzke <<a href="mailto:stepjak@yahoo.fr">stepjak@yahoo.fr</a>> escribió:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Hi Forest,<br>
<br>
Le vendredi 7 janvier 2022, 15:57:27 UTC−6, Forest Simmons <<a href="mailto:forest.simmons21@gmail.com" target="_blank" rel="noreferrer">forest.simmons21@gmail.com</a>> a écrit : <br>
> Very true, and I would have said err on the side of VSE until Robert B-J<br>
> convinced me that sincere cycles are practically non-existent with an occurrence<br>
> of less than 0.5 percent.<br>
> <br>
> He got that statistic from Fair Vote's analysis of over 400 elections, and they<br>
> have no particular reason to exaggerate that statistic. So suppose they're<br>
> right, then do we conclude that any old Condorcet method is as good as another?<br>
<br>
I would think that I wouldn't trust a study of IRV elections to talk about voted<br>
CWs across all methods, for the same reason I wouldn't trust a study of FPP<br>
elections to tell me that most elections will have a voted majority favorite<br>
across all methods. The incentives aren't the same.<br>
<br>
IRV has no truncation strategy. (In Burlington, we see truncation anyway; how<br>
much more would we see if the method could actually reward truncation?) If there<br>
is usually a *sincere* CW (which I think is likely), then it's easy to imagine<br>
that more complete rankings might lead to finding that CW.<br>
<br>
IRV also has among the highest compromise incentive, which I believe translates<br>
directly to nomination disincentive. Meaning, those 400 elections probably<br>
could have had more candidates nominated under a different method. Particularly<br>
if these were not very viable candidates, I think this could definitely lower<br>
the rate of seeing a voted CW.<br>
<br>
If we enact a Condorcet method and find that, indeed, there is almost always a<br>
voted Condorcet winner, I would be inclined to wonder if there is something<br>
wrong. Because given a couple of assumptions:<br>
<br>
1. Voters feel inclined to truncate the candidates they feel they are trying to<br>
defeat, or that they don't care to worry about for whatever reason.<br>
2. Candidates feel free to enter the race even if they can't win, and their<br>
supporters feel free to vote for them despite this possibility.<br>
<br>
Then, this should create an environment where voted cycles are certainly<br>
possible. In particular, cases where a less viable candidate manages to get a<br>
pairwise win over a more viable candidate, due to considerable abstentions of<br>
voters who either were not interested in that contest, or perceived it as unsafe<br>
to support the lesser evil between the two.<br>
<br>
Kevin<br>
<br>
</blockquote></div>